Emergency Evacuation Modelling: A Systematic Review on Human Behaviour Prediction Methodologies and Approaches for Planning the Evacuation Routeing

Abstract

Abstract Natural disasters are frequently occurring events, and evidence indicates that they will continue to increase in frequency and cause further harm and damage to society and the environment. Thus, robust and effective disaster management strategies are necessary to overcome the consequences of disastrous events. To enhance evacuation planning that considers the major risk mitigation aspect, this work aims to deliver an informative database of recent methodologies used in planning for evacuation in an attempt to bridge the gap between the social sciences and transportation engineering. A systematic review of 70 articles and conference proceedings in the field of planning for evacuation concerning behaviour prediction and the optimisation of evacuation routeing was carried out to offer broad insights into the improvement areas in the utilised methods. The findings revealed that agent-based simulation is the most employed method in investigating/predicting human behaviours during emergencies. Further, the traffic assignment optimisation approach is the most frequently used technique in planning/analysing evacuation routeing, whereas the objective of identifying the shortest path is commonly used. Further, the decision to evacuate is among the most investigated human factors in planning the evacuation process, as it directly affects evacuation demand and departure times, and hence, traffic management. After a detailed exploration of evacuation planning methodologies, this work proposes a framework comprising sub-models that transfer the outputs from human behaviour investigations into evacuation routeing planning, with suggestions regarding the suitable approaches that can be utilised. Extended research could be implemented to check the compatibility of the suggested methodologies across the framework.